Post-doctoral researcher in Natural Language Processing for Health Data Science

The position is embedded within the department of cardiology and primarily focusses at the development, clinical evaluation and implementation of natural language processing techniques, with a specific focus on the use of large language models (LLMs) for different tasks. Specifically, you will contribute to several running projects, focused at 1) LLM-based structuring of clinical notes for optimized data extraction, 2) evaluation of LLM-based guideline adherence assessment, and 3) development and testing of explainable AI-enabled clinical decision support tools. These projects are embedded in the EU Horizon DataTools4Heart, AI4HF and AIXPERT projects, which focus on the development of trustworthy and explainable AI tools in cardiology. Specifically, DataTools4Heart focusses on the development of tools for the re-use of electronic health care data for research purposes. AI4HF focusses on the development of trustworthy AI models for clinical implementation in heart failure patients and AIXPERT focusses on the development and evaluation of trustworthy LLM-based systems for clinical decision support tools.

Additionally, the successful candidate will directly collaborate with startup company HealthSage AI that developed their own Large Language Model for clinical workflow optimalisation. During the project, the LLM model will be optimized and evaluated in the context of different clinical workflows, including both out-patient and (acute) in-patient care settings.

We are looking for a post-doctoral candidate who is eager to design, implement, and evaluate responsible NLP methods in the context of the different projects. You will propose state-of-the-art LLMs and will develop Dutch and/or multilingual linguistic resources with a strong focus on healthcare applications in the field of cardiology. You will actively participate in the clinical evaluation and implementation of these models and work together with other post-doctoral researchers across Europe. You will have the opportunity to apply statistical and machine learning techniques to a range of clinically relevant problems in cardiology, which will be defined and discussed together with medical specialists.

The successful candidate will join a cross-disciplinary team consisting of computer scientists, medical professionals, PhD students, and post-doctoral researchers and will also have the opportunity to work together with informaticians and software engineers within Amsterdam UMC. The preferred background combines technical knowledge and strong clinical background, to be able to design and carry out clinical evaluation and implementation studies, supported by technical researchers. You will provide efficient and scalable implementations of your methods, and will integrate them with popular open-source systems.

We are looking for a highly motivated post-doctoral researcher with a strong technological and clinical background, preferably with knowledge of natural language processing systems.

Prerequisites:

• PhD degree in machine learning, artificial intelligence, medical informatics, statistics, data science, or related fields, combined with strong clinical background;
• Hands-on experience in natural language processing, machine learning, medical informatics and/or data science;
• Hands-on experience in clinical evaluation and implementation studies;
• Highly motivated with a creative and independent mindset;
• Programming skills in Python and experience implementing machine learning systems (e.g., scikit-learn, PyTorch);
• Capable of working in multidisciplinary environments;
• Good command of English, both spoken and written.

Not required, but helpful:

• Experience with prediction modelling, data harmonization and ontologies;
• Experience with clinical data and/or cardiology;
• Contribution to open-source projects.